TOWARD EARLY INTERVENTION: MODEL OF ACADEMIC PERFORMANCE IN A CDIO CURRICULUM

After three years of implementation of the CDIO initiative in the electronics engineering program at the Pontificia Universidad Javeriana, the curriculum management has focused the operation of the program on monitoring students who, from the point of view of the assessment of learning, generate important information to the program evaluation. The performance of the students is an important marker that indicates the efficiency of the program and represents the level of success of the reform according to the CDIO methodology. The structure of the curriculum and the gradualness of the integrated competences, reflect a program transition behavior that is aligned with the student development model of the university. Three transitions were identified: first year, second and third year, and advanced students. These transitions show different behaviors and needs that, in the institutional context of risk prevention, involve the identification of realities that require early monitoring and intervention. In order to implement, the student development model, the university generates a risk prevention program that takes into account individual, psychosocial, academic and financial factors. Based on this model, a system of early alerts is created. This system includes intervention and monitoring processes. The initiative is complemented by a student accompaniment program (PAE + N, by its initials in Spanish), which is being developed initially in the School of Engineering. Under this context, it is necessary to design and implement models to identify patterns associated with academic performance and transitions of undergraduate students. This project is developed with the aim of detecting problems which can be intervened by making use of the entire offer of accompaniment from the university (advisors, workshops, psychological counseling, etc.). These patterns are detected using variables available in the University's information ecosystem, using analytical techniques and artificial intelligence. In this paper, the identification methodology for risk patterns is shown. Additionally, some of the alerts that are in development are described including the analysis of their incidence as efficiency indicators of the CDIO program. The results of this project will allow reforms to the courses, the program, the teaching methodologies, learning and assessment, as well as the programs of the accompaniment of students in all transitions.